In early 2012, Cort Johnson began an interview with the Irish ME Association Officer and peer-review published author, Tom Kindlon. This interview project was revived and edited by Stukindawski in early 2013. The interview is in three parts, with an introduction by Stukindawski, and the other parts can be accessed by clicking the following links:-
In this first part of the interview, Cort and Tom discuss the definition of the term CBT. In the field of ME/CFS the term comprises a variety of distinct and multi-faceted treatment options. Tom highlights the differences between programs and terms while also taking a look at the beliefs about ME/CFS that the prescription of CBT is based on. Later, Cort and Tom discuss the differing patient responses to CBT. Cort asks whether there are predictors for a positive response such as severity and Tom guides us through the relevant literature.
CORT JOHNSON: CBT is a single term for a very broad therapeutic approach that includes pacing, sleep hygiene and thought management. Some of these are used by many patients and are considered acceptable ways to manage this disease (pacing, sleep hygiene) while others (thought management) stir up issues. Is there any way to tell how effective one part of CBT is from another?
Tom Kindlon: I see there is no easy pitch to start me off!
I want to first apologise in advance for the length of my responses. Hopefully those who stick with me will find something of interest.
Anyway, getting back to your question: as with a lot of things in the ME/CFS field, answering it isn’t that straightforward. You say that CBT includes pacing. However, traditionally, most CBT interventions for ME/CFS would have avoided such language, and associated themselves more with graded activity or graded exercise. “Pacing” has generally been more associated with “listening to one’s body” while most CBT interventions for ME/CFS are based on the “fear avoidance” model: “this theory regards chronic fatigue syndrome as being reversible and that cognitive responses (fear of engaging in activity) and behavioural responses (avoidance of activity) are linked and interact with physiological processes to perpetuate fatigue” (from the Lancet PACE Trial paper). As the PACE Trial manuals make clear, this form of CBT “consider[s] [an] increase symptoms as natural response to increased activity” and specifically says CBT (and GET) do not involve “encourag[ing] participants to listen to their body”. Interestingly, despite these descriptions, they still use the term “pacing” in the description of CBT: Complex Incremental Pacing /Cognitive Behavioural Therapy (CBT).
The use of the term “pacing” in different ways prompted Drs Ellen Goudsmit, Jo Nijs, Leonard Jason & Karen Wallman to recently issue a consensus statement on what pacing is. Their paper concentrates more on symptom-contingent pacing but also discuss programs combining time and symptom-contingent pacing. Symptom-contingent pacing means that somebody’s symptoms informs them how much to do and when to stop. This is in contrast to time-contingent strategies which mean that when people rest, or continue, is decided by a timer/the schedule and not the symptoms they feel. CBT programs tend to use more time-contingent pacing. Such programs tend to try to stabilise patients and try to avoid “boom-and-bust” by identifying patterns from activity diaries.
Incidentally, although a lot of focus of some CBT programs is on the graded activity/exercise element (rather than the stabilisation element), I have wondered if the reason some patients report some improvements is to do with such programs leading to patients stabilising their activities more than they might have done previously.
The ideal way of working out how effective different elements of a CBT program are would be to have trials that involve interventions that are only different in one way: the presence or absence of a particular element. However, generally what happens is that there are several differences between programs that are being compared. Meta-analyses can also try to look at such issues. A review by Malouff and colleagues in 2008 divided up programs which had had a graded activity element, into those that had or did not have a cognitive element i.e. where “unhelpful thoughts” are explored in therapy and a patient is taught to challenge them. It found that programs that didn’t have a cognitive element had better results but the difference was not statistically significant.
Cognitive behavior therapy (CBT) programs for ME/CFS are often based on a model of chronic fatigue syndrome that posits that fatigue and functional impairment are perpetuated by physical inactivity, somatic attributions, focusing on bodily symptoms and a low sense of control. Mediation analysis has been used to explore this model. A paper by Wiborg and colleagues found that changes in physical activity were not related to changes in fatigue scores. Drs. Stouten and Goudsmit subsequently analysed the Prins et al Lancet paper (2001) which compared CBT, guided support and treatment as usual. They found that “CBT had no significant impact on somatic attributions and focusing on bodily symptoms, and that in line with established guidelines, these two variables were not mediating factors. The only variable in the model showing an effect of CBT was sense of control.” This was an interesting finding.
A London study investigated the “active ingredients” of CBT and counselling in patients with unexplained chronic fatigue (for greater than three months) using principal component analysis (PCA) of possible predictive factors. It found that there was some overlap between the treatments. Looking at the results on the fatigue scale, the key predictor of a good outcome was “emotional processing, including the expression, acknowledgement and acceptance of emotional distress.”
My preference would be for a greater focus on objective measures in the ME/CFS field particularly with non-pharmacological interventions. Subjective measures are potentially subject to a whole range of response biases, meaning that one can’t necessarily have full confidence in the results. It is human nature, for example, to want to believe the effort put into a therapy has led to an improvement and also to want to please, or not displease, a therapist who has worked with you closely and whose opinion of you may potentially be influential in other settings, including social and medico-legal. Also, if different therapies encourage one to focus more or less on one’s symptoms, two patients may report the magnitude of the same level of symptoms in different ways. This leads me to be cautious about interpreting the results of a lot of the trials in the field.
Getting back to your initial question, I don’t think attempts have been made to evaluate sleep hygiene advice. So my comments on it are more informed by observation rather than any specific research findings.
While “sleep hygiene” suggestions can all seem reasonable on the surface, just like with a lot of advice in the field, some aspects could be problematic for some individuals with ME/CFS. So, for example, advice to not take caffeine later in the day seems sensible, as does having a quiet place to sleep, a room free of as much extraneous light as possible. However, is napping during the day necessarily bad? This appears to be an idea that has borrowed from healthy people or other groups of patients who don’t have the rest needs – or the post-exertional complex of symptoms – that people with ME/CFS have.
Similarly, sleep hygiene regimes may recommend getting up at in or around the same time every day. However, if for one reason or another, somebody with ME/CFS hasn’t got many hours of sleep at that stage, many would see it as a bit risky for the patient to necessarily commit to getting up at the same time on such a day – doing that too many times might trigger a flare-up or relapse from the patient having insufficient rest. Hopefully more research will be done in the future to discover what causes the sorts of sleep problems associated with ME/CFS (e.g. hypersomnia, insomnia, sleep reversal, etc.) helping to inform “sleep hygiene” programs.
CORT: Like every therapy for CFS some people appear to benefit from some forms of CBT while others do not. I believe a Jason study suggested that low cortisol levels in people who were presumably sicker could impair them from benefiting from these types of therapies. If I remember correctly, another study, on the other hand, suggested these therapies could help to normalize cortisol levels. It seems possible that illness severity might play a role with people who are less ill benefiting more. Do researchers have any idea who tends to benefit and who does not?
TOM: It is increasingly recognised and accepted by “all sides” that there are a variety of individuals being diagnosed with “Chronic Fatigue Syndrome” i.e. that the condition is heterogeneous. Yet, research looking at subgrouping patients (in some way) is underdeveloped. This is disappointing.
It is interesting to look at the Fukuda et al. (1994) criteria paper – probably the most cited paper in the field: the authors suggested that future studies (i.e., diagnostic, treatment, epidemiological, basic science, etc.) subgroup or stratify patients by severity, presentation, lab or other diagnostic test, co-morbidities, etc. As I mentioned, most studies have not done this and even when it is done, certain areas, such as depression, have been emphasized while others have been barely touched. I highlighted this issue in a letter which was published in Psychological Medicine, “Stratification using biological factors should be performed in more CFS studies” .
The study by Dr. Leonard Jason and his team in DePaul University, Chicago you refer to actually combined the data from the four separate arms of a trial: the interventions tested were the graded activity form of CBT used in King’s College London, a form of exercise therapy, a cognitive therapy program involving pacing devised by Dr. Friedberg, and a relaxation program. I think they combined the data as the sample sizes were small (less than 30 for each intervention). This study involved outpatient appointments so patients who were bedbound, housebound or used wheelchairs didn’t take part. In one paper, they concentrated on cortisol levels. As you know, in healthy individuals, cortisol has a daily, or “diurnal”, rhythm – the levels aren’t static across the day. So in this study, they defined cortisol levels as abnormal “if they continued to rise, were flat, or were abnormally low over time” (they used five time points), as opposed to simply a low total level. They found across the treatments, the participants with abnormal cortisol before starting treatment appeared not to improve over time, whereas those with normal baseline cortisol showed improvements in a number of immunologic and self-report measures. In another paper on the same group of patients, they concentrated on results from the SF-36 physical functioning questionnaire. If an individual increased by 10 or more on this scale, they were considered to have shown positive change. They found that a variety of measurements taken before therapy started differentiated patients who showed positive change, including immune function results, time logs, sleep status and past psychiatric diagnosis [those who improved were more likely to have had a past psychiatric diagnosis (74% vs 48%) although current psychiatric status as measured by a SCID assessment was not statistically different at baseline].
I’m guessing the study you are referring to, which found that cortisol increased on average following CBT, is one by the King’s College London team. This was an uncontrolled study and the authors themselves say that they couldn’t control for other factors such as illness progression. They also say that the diurnal pattern of cortisol release was not altered by CBT. To complicate matters, the same five authors were part of a six-person team who reported that, in another group of CFS patients, a “flattened diurnal profile of salivary cortisol was associated with a poor response to CBT”. They also found that when they analysed the free cortisol levels in urine collected for 24 hours (at baseline), lower levels were associated with a poor response to CBT in those who were not taking psychotropic medication. They concluded that low cortisol is of “clinical relevance in CFS, as it is associated with a poorer response to CBT”. So the situation is complicated.
I found one other possible predictor when trawling through the literature: back in 1991, a King’s College London team (again involving Drs. Chalder & Wessely) reported that there was a trend (p=0.08) for those with a positive VP1 test to do worse with CBT. VP1 is a test for a specific enteroviral antigen; it has generally been out of favour in ME/CFS circles for a while, but I believe that Dr. John Chia thinks it has a value. So that finding is intriguing.
Another way to categorise patients are written criteria for diagnosis. Many ME and CFS advocates in the UK have said for years that the reason many of the trials there have been relatively successful is that the trials only required participants to satisfy the Oxford criteria – basically unexplained physical and mental fatigue that has a debilitating effect, is not life-long and has lasted six months or more. This means participants do not need to have other symptoms commonly associated with the ME complex. A London trial of graded exercise or cognitive behaviour therapy looked at those with chronic fatigue (which had lasted at least three months) – so a cohort not that different to an Oxford criteria CFS group. It found that meeting the Fukuda criteria for Chronic Fatigue Syndrome was the most powerful predictor of a poor outcome. The recently published PACE Trial did not report such an effect when looking at two primary outcome measures. However, some have raised questions about the criteria used including how the criteria were operationalised and the exclusions used. Only 340 of the 640 participants did not have a psychiatric disorder and hence were eligible (according to Prof. White) to satisfy the ME criteria used (the London criteria). Yet, there were said to be 329 participants with ME in the trial. In other words, 97% (329/340) of those with Oxford criteria without psychiatric disorders were said to have ME, an extremely high percentage. It’s important to note that the PACE Trial analyses in the Lancet were only regarding average results – they don’t tell us exactly who might be report an adverse event.
One recent paper, again examining the effects of CBT in King’s College London, divided the CFS patients up into four groups based on their response. The researchers reported that a poor response was characterised by more frequent weight fluctuation, physical shakiness and pain, and higher anxiety and symptom focusing scores compared to the other classes. The research team actually looked at 38 baseline factors but we were not given data on all factors so I am left wondering whether there were other factors that also categorised this group.